3D Image Volumes From 2D Digitally Reconstructed X-Rays: A Deep Learning Approach In Lower Limb CT-Scans.

PURPOSE Three-dimensional reconstructions of the human anatomy have been available for surgery planning or diagnostic purposes for a few years now. The different image modalities usually rely on several consecutive two-dimensional acquisitions in order to reconstruct the three-dimensional volume. Hence, such acquisitions are expensive, time-demanding and often expose the patient to an undesirable amount of radiation. For such reasons, along the most recent years, several studies have been proposed that extrapolate three-dimensional anatomical features from merely two-dimensional exams such as X-rays for implant templating in total knee or hip arthroplasties. METHOD The presented study shows an adaptation of a deep-learning-based convolutional neural network to reconstruct three-dimensional volumes from a mere two-dimensional digitally reconstructed radiograph from one of the most extensive lower limb computed tomography datasets available. This novel approach is based on an encoder-decoder architecture with skip connections and a multidimensional Gaussian filter as data-augmentation technique. RESULTS The results achieved promising values when compared against the ground truth volumes, quantitatively yielding an average of 0.77 ± 0.05 structured similarity index. CONCLUSIONS A novel deep learning based approach to reconstruct three-dimensional medical image volumes from a single x-ray image was shown in the present study. The network architecture was validated against the original scans presenting SSIM values of 0.77 ± 0.05 and 0.78 ± 0.06, respectively for the knee and the hip crop.